Jagpreet Singh Mittu:
Building the Future of Semantic Intelligence.
AI Researcher & Software Engineer specializing in Llama 3.1, RAG pipelines, and Voice-Driven Robotics.
Bridging Data Engineering and Generative AI.
Based in Dallas, TX, and finishing my B.S. in Computer Science and Engineering at the University of Toledo — graduating May 2026.
I bridge the gap between Data Engineering and Generative AI. My research focuses on optimizing model latency and improving candidate-relevance algorithms using Hybrid AI systems and Llama 3.1.
Certifications · Be10X — Power BI · Coursiv — Generative AI
A Chronological Build Log.
AI & Machine Learning Intern
- ▹Designed a Three-Tier Ranking Algorithm on spaCy 300D vectors — +40% relevance over keyword search.
- ▹Built an SVM resume classifier and a semantic matching engine over 2,400+ resumes.
- ▹Integrated Llama 3.1 (Groq API) for context-aware candidate assessments and gap analysis.
- ▹Optimized the Streamlit dashboard with caching — load times from ~30s down to <1s.
Technical Pre-Sales Specialist
- ▹Led Proof-of-Concept projects on cybersecurity and network technologies.
- ▹Performed technical pre-qualification of prospects across cloud & security stacks.
- ▹Translated client requirements into actionable technical recommendations.
Data Engineering Intern
- ▹Optimized ETL workflows and indexing strategies — 10% query performance improvement.
- ▹Built and maintained robust data pipelines using Argus Agent for monitoring.
- ▹Tuned database systems to enhance storage and retrieval efficiency.
Selected Research & Builds.
Hybrid AI Talent Intelligence Platform
A semantic talent engine built during my Sedha Consulting internship. A Three-Tier Ranking Algorithm fuses cosine similarity over spaCy 300D word vectors, keyword bonuses, and categorical boosts — yielding a 40% relevance boost over keyword search. Streamlit caching cut dashboard load times from ~30s to <1s.
AAP.JS — Voice Interaction System
Voice-driven assistant on a Raspberry Pi 4 stack, fusing Azure Cognitive Speech-to-Text (97.94% accuracy) with the DeepSeek API for context-aware responses. Sustained 3+ hours of continuous, stable public interaction during the Senior Design Expo.
GenAI Chatbot Ecosystem
LLM chatbot with function calling (Google Books), DALL·E 3 image generation via Azure OpenAI, and a Pinecone + LangChain memory stack. Deployed on Azure Static Sites with 20% lower latency.
A note from the field.
“One of Jagpreet's key strengths is his ability to translate theoretical knowledge into working systems. He went beyond basic keyword-based approaches and designed a more meaningful, context-aware matching system, integrating Llama 3.1 for candidate assessments and skill-gap analysis. I recommend him without hesitation.”
The Technical Stack.
Let's build something intelligent.
Open to AI Research, ML Engineering, and Software Engineering roles starting mid-2026. Reach out for collaborations, research, or a coffee chat.